Abstract
Objectives: Longus colli tendinitis (LCT) is a rare, self-limiting disease primarily characterized by neck pain. This study is to investigate and analyze the imaging and clinical features of LCT and to develop a predictive model for pain risk in LCT based on these features. Methods: This study included 35 patients with LCT enrolled between January 2017 and December 2024. Radiological features, laboratory indicators, and clinical profiles were systematically analyzed. We stratified LCT patients into high-risk (n = 20) and low-risk (n = 15) groups based on pain intensity and duration. Nomograms were developed using logistic regression models, with feature selection performed via the least absolute shrinkage and selection operator method. Model performance was evaluated through discrimination (Harrell's C-index) and calibration (calibration plots), with internal validation conducted via bootstrapping. A clinical impact curve was used to assess the model's clinical usefulness. Results: MRI features of LCT included average lesion width of 6.13 mm, length of 64.00 mm, circumference of 134.52 mm, and area of 230.64 mm(2). Clinically, LCT patients exhibited elevated white blood cell counts, neutrophil counts, hsCRP levels, and IL-6 levels. Feature selection revealed that the lesion area could predict pain risk in LCT patients, which was used to construct a predictive model. The model demonstrated a C-index of 0.93 (95% CI 0.84-0.99). Internal validation confirmed the model's robust performance, with a C-index of 0.93 (95% CI 0.83-0.99). Conclusion: LCT possesses distinct imaging and clinical features. Utilizing these features enables effective prediction of pain risk, thereby assisting clinical decision-making.